1. Field of the Invention
This invention relates to the field of ultra-fast laser operation and the characterization of optical pulses.
2. Background of the Invention
Within the research field of ultrashort-pulse lasers and more generally within ultrafast science and technology, there is a need to develop and implement a range of diagnostic techniques for characterizing ultrashort optical pulses that have durations of femtoseconds (fs). At limits where pulses can be shorter than 10 fs, the diagnostics used for quantitative assessments must exhibit a time resolution in the region of ifs. Indeed, driving the development for better characterization has been the continued development of ultra-fast lasers with increasing gains in power and portability. For example, driven by the development of all optical networks with increasing bandwidth, mode-locked erbium doped fiber lasers that emit sub-picosecond pulses at the telecommunication wavelengths have been demonstrated with average powers of several hundred mW. See for example Hofer et al. in Opt. Lett. 23, 1840 (1998), the entire contents of which are incorporated herein by reference.
Characterizations of femtosecond and picosecond pulses in the ultra-fast lasers in the past have involved techniques based on second-order nonlinear processes using non-collinear or collinear geometries providing intensity and interferometric autocorrelations, respectively. Such techniques have been reported by Diels et al. in Appl. Opt. 24, 1270 (1985), the entire contents of which are incorporated herein by reference. Other second-order autocorrelation techniques have been based on two-photon absorption processes in semiconductors. See for example Tagaki et al. in Opt. Lett. 17, 658 (1992), the entire contents of which are incorporated herein by reference. Second order nonlinear processes are usually sufficiently sensitive so that low energy pulses can be characterized. However, second order nonlinear processes suffer from limited sensitivity and are specific only to particular pulse shapes. For instance, second-order autocorrelations when invoking symmetrical functions do not provide information on asymmetric pulses. Furthermore, since second-harmonic signals are typically generated in thick crystals in a geometry that satisfies phase-matching conditions, second order nonlinear process are polarization sensitive and have a limited wavelength tunability range.
Besides second order nonlinear processing, third-order nonlinear processes have been recognized as a method for characterizing ultra-fast laser pulses. In principle, third-order nonlinear processes are superior to second-order ones. However, third-order nonlinear processes are generally limited in sensitivity due to the lack of materials with both a strong third-order nonlinearity and transparency. Third-order techniques prior to the present invention have typically been implemented with fused silica as the optical material. More specifically, as described by Streltsov et al. in Appl. Phys. Lett. 75, 3778 (1999) and by Langlois et al. in Opt. Lett. 24, 1868, (1999), the entire contents of which are incorporated herein by reference, third-order autocorrelation can be implemented by using three-photon absorption in photodiodes. The sensitivity of a third-order autocorrelation using three-photon absorption is somewhat limited. Further, a simultaneous measurement of the spectrum of the nonlinear signal is required to retrieve information on the phase of the pulse. Nonetheless, third-order autocorrelation techniques offer advantages over second-order techniques. See for example Meshulach et al. in J. Opt. Soc. Am. B 14, 2122 (1997), the entire contents of which are incorporated herein by reference.
Solutions based on three-photon absorption in photodiodes, while characterizing asymmetric pulses and removing direction-of-time ambiguity, only provide information on a laser pulse amplitude and not the phase. For the complete characterization of ultra-fast short pulses, full knowledge of both amplitude and phase is required. One technique for simultaneously retrieving information on the phase of the pulse is a frequency-resolved optical gating (FROG) techniques reported by Trebino et al. in Rev. Scientific Instruments, 68, 3277 (1997) and in U.S. Pat. No. 5,530,544, the entire contents of which are incorporated herein by reference. FROG is one technique, where through time-frequency domain measurements, both amplitude and phase information of optical pulses are derived. However, such pulse measurement and methods conventionally have required spectrometers coupled to expensive highly sensitive detector arrays.